Take a partner-led consultancy, three years from a planned sale. The founder is weighing what to do with the firm’s AI programme. The obvious move is to give the rainmakers better tools, let them turn work around faster, win more. The less obvious question is what that choice does to the firm’s valuation when a buyer runs the numbers and finds that the revenue walks out the door with two named partners.
What does a buyer of a professional services firm actually pay for?
A buyer of an accountancy, law firm, or consultancy is paying for the probability that revenue continues after the principals leave. The three things that determine this are client relationships held by the firm rather than by named partners, a delivery method the team can execute without those partners in the room, and a bench capable of covering significant departures.
The valuation gap between a founder-dependent professional services firm and a founder-independent one is material and well-documented. Lower-middle-market M&A research puts the discount for businesses with extreme key-person concentration at 30 to 50 per cent of market comparables. In EBITDA multiple terms, a well-run consulting practice with a professional leadership team and transferable client relationships trades at 7 to 8 times. The same practice, where the founding partners personally hold every major client relationship and approve every deliverable, trades at 3 to 4 times.
In professional services, this concentration has a particular shape. The clients follow named partners, not the brand. The delivery methodology lives in senior minds rather than in a documented framework. Work is won and retained through specific individuals showing up. Buyers price all of this risk in, and the formal valuation methodology backs it up: a firm of chartered accountants applying the key-person risk framework typically applies a discount of 10 to 25 per cent before any other adjustments.
Why does AI choice either lift or hurt your sale multiple?
There is a fork in every professional services AI programme, and many firms take the wrong turn. Giving the senior partner faster research tools, letting the lead solicitor draft precedents at double speed, these are real productivity wins. The catch is that they deepen key-person dependency, making those individuals harder to replace and the firm more reliant on their continued presence when a buyer arrives.
The alternative is to point AI at the firm’s knowledge, its methods, its precedents, its delivery frameworks, and use it to make that knowledge accessible to the whole team. Harvard Business Review has described the emerging structure in AI-native consulting firms as an obelisk rather than a traditional pyramid: fewer layers, with AI handling the research and analysis tasks that juniors previously carried, and senior roles shifting to engagement architects and client leaders rather than doing junior work at senior rates.
McKinsey’s deployment of this logic internally with Lilli, its internal knowledge synthesis system, shows what this looks like at scale. By 2024, over 70 per cent of McKinsey’s 45,000 employees were using it, averaging 17 queries per week, materially reducing research and planning time across the firm. McKinsey’s approach moved knowledge out of individual heads and into a system the whole firm could search. That same logic, applied at a 30-person consultancy or law practice, is precisely what a buyer’s due diligence team wants to see evidenced.
Where does exit-readiness AI actually show up in practice?
The use cases that matter for exit readiness are those that move knowledge from individual heads into firm-accessible systems. An AI tool that assists one expert is a productivity gain. An AI knowledge base that makes the firm’s expertise searchable for everyone on the team is a valuation asset. The two look similar from the outside but perform very differently in a data room.
In a legal practice, the concrete version is an AI-indexed precedent library that lets any fee-earner find the relevant prior work without asking the senior partner. In a consultancy, it is a delivery framework built from the firm’s actual project history, searchable and usable without the founding partner in the room. In an accountancy practice, it is client-specific service maps that document what each client needs, when, and how the firm has historically handled their work.
The AI mandate also functions as the forcing function that many firms have needed for years but never got around to. Writing down how the firm actually does the work, its playbooks, its house style, its decision criteria at each stage, is the hard work that generates the knowledge base. The SRA, for instance, requires solicitors to have documented governance frameworks covering technology use, leadership oversight, and quality monitoring. Meeting that requirement while building the firm’s knowledge base addresses two problems in one programme.
When should you start and what can wait until later?
The two-year window before a professional services firm sale is when this work has to happen. A buyer’s due diligence team will look for documented delivery methodology, client intelligence held in firm systems rather than in named partners, and evidence that quality standards operate without specific individuals in the room. A six-month sprint before exchange of contracts cannot build any of that from scratch.
Exit planning disciplines typically structure the preparation timeline at 24 to 36 months. For professional services, the knowledge codification work sits in the first 12 months: identify the highest-dependency areas, start pulling working methods out of senior heads and into documented frameworks, build the initial knowledge base from existing documents, precedents, and past deliverables. The AI tools accelerate this considerably, but they need the raw material first.
What can wait is the choice of specific products and integrations. The market moves fast, and the tools available now will have evolved significantly in 12 months. The foundational work of documenting methodology, organising historical knowledge, and building the delivery framework is what matters and what compounds over time. This is worth being clear about internally, because the conversation about AI often gets pulled toward product decisions before the underlying knowledge work has been done.
Lower-middle-market transaction data indicates that businesses with documented standard operating procedures receive materially higher valuations than those without, in some cases 20 to 40 per cent higher. For professional services firms, the equivalent is documented methodology. It does not have to be exhaustive; it has to be sufficient for a buyer’s team to satisfy themselves that the work can continue.
What else sits alongside AI in a firm that’s ready to sell?
AI handles the knowledge layer well. A professional services firm that commands a full multiple also needs client relationships held by the practice rather than by named individuals, and delivery work that a team leader can scope and deliver without the founding partner in every meeting. Both require deliberate structure alongside the AI programme; neither follows from the AI work alone.
The client relationship piece requires a CRM that actually holds client intelligence rather than one that sits empty while relationships live in email inboxes. It requires explicit account assignments where each major client has a named relationship owner other than the founding partner. And it requires those team members to be the primary contact, taking briefing calls, owning the quarterly review, being the person the client rings when something goes wrong.
The delivery piece requires that work is priced and scoped in a way that a team leader can commit to without the founding partner validating every judgment call. Fixed-fee or retainer models help with this because they force the firm to define scope clearly enough that someone other than the founding partner can deliver it consistently.
Half of UK SMEs are now using some form of AI, according to the British Chambers of Commerce. In professional services, firms heading for a sale need to channel that adoption deliberately. AI adoption without the underlying relationship and delivery structure in place is productivity without exit readiness.
The firm that arrives in a data room with its methods documented, its client intelligence systematised, and its delivery quality demonstrable across the whole team commands a meaningfully different multiple than one where the same revenue depends on two or three named partners. The AI programme that gets you there does not start with the most powerful tool. It starts with the decision to stop keeping the firm’s knowledge in partner-shaped containers.
